KR20210025015A - Information processing device, information processing system, and program - Google Patents

Abstract

It is estimated when an event occurs that will lead to damage to crops caused by pests. Using the acquisition unit, which acquires pest occurrence information including information at the time of damage to the crop caused by pests and environmental information including the cultivation environment of the crop, and the pest occurrence information, the damage to the crop occurs. An accident period determination unit that determines an estimated event occurrence period suspected of having a subsequent event occurring and a non-death period in which the event leading to the occurrence of the damage has not occurred, the environmental information of the estimated event occurrence period, and the nothing. There is provided an information processing apparatus including an event estimation unit, which compares the environmental information in an event period and estimates the occurrence time of the event in which the event occurs from the estimated event occurrence period.

Description

Information processing device, information processing system, and program

The present invention relates to an information processing device, an information processing system, and a program.

The occurrence of pests is a factor that influences the crop harvest. In recent years, research and development of technology for protecting crops from such diseases and pests using information and communication technology has been developed. In pests, there are three stages of infection, incubation, and onset (occurrence) of crops, and by predicting the occurrence of pests, various technologies that can help in efficient production of crops are being developed. In Patent Document 1 and Patent Document 2, techniques for predicting the occurrence of pests are disclosed. These are technologies that search for conditions leading to the occurrence of pests from past environmental information, crop cultivation information, and the like. Based on this prediction of the occurrence of pests, farmers are protecting crops from pests by spraying pesticides or the like.

Patent Document 1: Japanese Unexamined Patent Application Publication No. Hei 11-287871 Patent Document 2: Japanese Unexamined Patent Application Publication No. 2003-167975 Patent Document 3: Japanese Unexamined Patent Application Publication No. 2009-106261

However, spraying of pesticides is more effective than the onset (occurrence) of the pests if they are performed before or after infection with the pests. For example, Patent Document 3 discloses a technique for predicting a disease that may cause infection and selecting an optimal pesticide. In predicting infection of a disease as described above, it is important to grasp the relationship between infection and outbreak. On the other hand, in patent document 3, there is no specific description regarding the estimation of infection of a disease. It is difficult to detect infection like an onset, and it is not easy to estimate the time of infection from the time of onset and predict future infection. In particular, in the data before the outbreak, conditions leading to infection of the disease and conditions that do not lead to infection of the disease are mixed, and among them, it is difficult to estimate the time of infection. For the above reasons, in protecting crops from pests and pests, there has been a demand for an improvement in the precision of infection estimation for pests.

Accordingly, the present invention has been made in view of the above problems, and an object of the present invention is to be able to estimate the occurrence of events leading to the occurrence of damage to crops caused by pests, new and improved. It is to provide an information processing device.

In order to solve the above problem, according to an aspect of the present invention, an acquisition unit for acquiring pest occurrence information including information when damage to a crop occurs due to pests and environmental information including the cultivation environment of the crop, and the Using the information on the occurrence of pests, the death period determination unit determines the estimated event occurrence period suspected of having an event leading to the occurrence of the damage to the crop and a period of no death in which the event leading to the damage occurrence has not occurred. And, by comparing the environmental information of the estimated event occurrence period with the environmental information of the non-event period, and estimating the occurrence time of the event in which the event occurred from the estimated event occurrence period, an information processing apparatus comprising an event estimating unit. Is provided.

In addition, in order to solve the above problem, according to another aspect of the present invention, a computer obtains pest occurrence information including information at the time of damage to a crop caused by pests and environmental information including the cultivation environment of the crop, Using the acquisition unit and the pest occurrence information, to determine a period of occurrence of an estimated event that is suspected of having an event leading to the occurrence of the damage to the crop and a period of no death in which the event leading to the occurrence of the damage has not occurred. , An event period determining unit, by comparing the environmental information of the estimated event occurrence period with the environment information of the non-event period, and estimating the occurrence time of the event in which the event occurred from the estimated event occurrence period, as an event estimating unit A program to function is provided.

In addition, in order to solve the above problem, according to another aspect of the present invention, an acquisition unit for acquiring pest occurrence information including information at the time of damage to a crop caused by pests and environmental information including the cultivation environment of the crop, and , Using the information on the occurrence of the pests, the estimated event occurrence period that is suspected of having an event leading to the occurrence of the damage to the crop and a period of no death in which the event leading to the occurrence of the damage is not occurring, is determined. Information comprising a determination unit, an event estimation unit, which compares the environmental information of the estimated event occurrence period with the environment information of the non-event period, and estimates a time when the event occurs from the estimated event occurrence period. A treatment system is provided.

According to the above configuration, environmental information of the estimated event occurrence period including at least one period of occurrence of an event leading to damage to the crop, and environmental information during an undead period in which no event leading to damage to the crop has occurred. By comparison, it is possible to more accurately estimate the event occurrence time from the estimated event occurrence period.

As described above, according to the present invention, it is possible to more accurately estimate the occurrence of events leading to damage to crops caused by pests.

1 is a conceptual diagram showing a concept according to an embodiment of the present invention.
2 is a conceptual diagram of machine learning used in the embodiment.
3 is a block diagram showing functions and configurations of the information processing system according to the embodiment.
4 is a diagram showing an example in which an event occurrence is estimated in the information processing system according to the embodiment.
5 is a flowchart showing an example of an operation flow of the information processing system according to the embodiment.
6 is a flowchart showing an example of an operation flow of the information processing system according to the embodiment.
7 is a flowchart showing an example of an operation flow of the information processing system according to the embodiment.
8 is a diagram showing an example of a hardware configuration of the information processing system according to the embodiment.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. In the present specification and drawings, elements having substantially the same functional configuration are denoted by the same reference numerals, and redundant descriptions are omitted.

<1. Outline of information processing system>

(1.1. Technical overview)

The present invention relates to a technique for estimating the occurrence of an event that will lead to damage to a crop, based on information on the occurrence of damage caused by a pest on a crop and the environmental information of the crop.

Referring to Fig. 1, an information processing system 1 including an information processing apparatus 100 for estimating the occurrence time of the event and other configurations for acquiring and outputting information necessary for estimating the occurrence of the event will be described. The information processing system 1 includes an information processing device 100, a cultivation sensor 300, a weather server 400, and an information processing terminal 200.

When a pest occurs in a crop, the information processing device 100 acquires information about the occurrence of a pest, such as input by a user or the like, to the information processing terminal 200. The information on the occurrence of pests and pests here is information including, for example, information when a pest occurs in a crop. Based on this information on the occurrence of pests and the environment information obtained from the cultivation sensor 300 and the meteorological server 400 for acquiring information on cultivation in the house, the information processing device 100 provides an event that will lead to damage to crops. The time of occurrence of this event is estimated.

The information processing apparatus 100 uses a machine learning technique called Multiple Instances Learning when estimating the occurrence of this event. By using this method, among events that will lead to damage to crops mixed in environmental information and events that will not be connected, feature quantities that will lead to damage to crops are extracted, and the probability of occurrence of events is calculated based on the feature amounts. And, based on the probability, the occurrence time of the event is accurately estimated.

(1.2. Machine Learning)

With reference to FIG. 2, a machine learning method using multiple instances learning will be described. In this method, for example, when a plurality of samples are classified into two, it is possible to grasp by what the classification is classified. In this method, a plurality of samples are handled as one bag. Here, description will be given on the case of classifying a plurality of samples into two, but the classification is not limited to two and may be performed in a plurality of three or more.

In FIG. 2, scenery 11, scenery 12, scenery 13, scenery 14, scenery 15, scenery 16, and scenery 17 of seven types of rooms are shown. Each landscape includes various items S, item X1, item X2, item X3, item X4, and item X5.

The landscape 11 includes the item S and the item X1, and the landscape 12 includes the item X1, the item X2, and the item S. Landscape 13 includes item X2, item X3, and item S, and landscape 14 includes item X2, item S, and item X4.

The landscape 15 includes the item X5 and the item X6, the landscape 16 includes the item X1 and the item X5, and the landscape 17 includes the item X4 and the item X5.

When such scenery 11 to 17 are taken as a plurality of samples, in this method, first, scenery 11 to 14 are set as one Bag, and scenery 15 to 17 are set as another Bag. Specifically, scenery 11 to 14 are determined as PositiveBag (hereinafter referred to as PB). On the other hand, scenery 15 to 17 are determined as NegativeBag (hereinafter, referred to as NB).

Here, in PB, at least one factor for classifying the scenery 15 to the scenery 17 and the scenery 11 to the scenery 14 is included. On the other hand, in NB, none of the factors for classifying landscapes 15 to 17 and landscapes 11 to 14 are included. In this way, bags of PB and NB are determined from a plurality of samples.

In this method, factors in each Bag that have an influence on the classification of PB and NB are extracted by comparing the Bag classified as PB and Bag classified as NB from a plurality of Bags classified as PB and NB as described above.

In Fig. 2, by comparing Bags classified as PB and NB, it can be found that PB and NB are classified as whether or not item S is included. That is, when comparing the scenery 11 to 14 and the scenery 15 to 17, the items S are included in the scenery 11 to 14, but the item S is not included in the scenery 15 to 17. Here, it can be seen that the scenery 11 to 14 are classified as PB, and the scenery 15 to 17 are classified as NB, with the item S as a feature quantity as an influence factor. Thus, in this method, by comparing PB and NB, it was found that the characteristic amount at the time of classification was S.

In the present technology, when a positive sample including a feature amount and a negative sample not including a feature amount are mixed in the PB, a feature amount dividing the PB and the NB can be found. Further, in this method, by discovering the characteristic quantity as described above, it is possible to determine whether the unknown Bag is positive or negative. In addition, in this method, it is possible to determine which bag in PB is a positive sample or the like by using the characteristic amount.

Using the above technique, in the information processing apparatus 100 according to an embodiment of the present invention, when an event leading to damage to crops is suspected to occur, environmental information of the occurrence period is PB, and the event occurs. Compare with environmental information NB during the uneventful period. By this comparison, the feature quantity leading to the occurrence of the event is grasped, and the occurrence time of the event is estimated from the period of occurrence of the estimated event time.

In addition, in the present invention, machine learning of an event occurrence prediction model based on estimation at the time of an event occurrence is performed. As described above, by machine learning the event occurrence prediction model, it is possible to autonomously update the event occurrence prediction model. Therefore, it is possible to improve the accuracy of calculation of the probability of occurrence of an event.

In addition, the information processing apparatus 100 newly inputs environmental information of crops, specifically, a predicted value of environmental information about a crop, or a current value, to the event occurrence prediction model after machine learning. Based on such environmental information, the information processing apparatus 100 may output a change over time in the probability of occurrence of events in the present or in the future with respect to the crop. In addition, it is not limited to the present or the future, and it goes without saying that the probability of occurrence of an event may be calculated for the past as well.

<2. Function configuration of information processing device>

The outline of an information processing system according to an embodiment of the present invention has been described above. Next, a functional configuration of the information processing system 1 according to an embodiment of the present invention will be described. As a specific example, an example of estimating the occurrence of an event leading to the damage will be described when a damage caused by a pest is caused to a crop. Hereinafter, the event will be described as infection with a pest and occurrence of damage as an outbreak.

Referring to Fig. 3, the functional configuration of the information processing system 1 will be described. 3 is a block diagram showing an example of a schematic functional configuration of the information processing system 1 according to the embodiment of the present invention.

As shown in FIG. 3, the information processing system 1 includes an information processing device 100, an information processing terminal 200, a cultivation sensor 300, and a weather server 400. In the information processing system 1, the information processing device 100 uses the information acquired from the information processing terminal 200 and the cultivation sensor 300 or the meteorological server 400 to predict the infection of a crop. Do. Hereinafter, first, the information processing terminal 200 for acquiring information required when the information processing apparatus 100 performs information processing, the cultivation sensor 300, and the meteorological server 400 will be described in order.

(Information processing terminal)

The information processing terminal 200 includes an input unit 210, a display unit 230, and a communication unit 220. The information processing terminal 200 has a function of acquiring disease and pest occurrence information, and the acquired pest occurrence information is output to the acquisition unit 110 of the information processing apparatus 100. The information processing terminal 200 may be a portable terminal such as a smartphone or an information processing device such as a computer. In addition, the pest occurrence information does not necessarily need to be acquired through the information processing terminal 200, and the acquisition method is not limited as long as it can be acquired by the information processing device 100.

((Input part))

In the input unit 210, pest occurrence information including information when a disease caused by a pest or the like has onset on a crop is input by input by a user or the like. The information on the occurrence of pests includes information on the date and time of the onset when a pest or the like caused by a pest has onset on a crop. Onset refers to detection or recognition that damage has occurred to the crop or that the incubation period has ended after infection, and symptoms of infection due to pests have developed. For example, the determination of the outbreak may be made based on information detected by the sensor or the like, if there is a sensor for observing the growth status of the crop, and the user recognizes the outbreak on the crop when observing the crop. It may be done by doing. Incidentally, the onset time represents an arbitrary period including the time when it is judged to be onset, and may be an onset date or an onset time.

The information on the occurrence of pests includes information on the types of pests that have occurred, in addition to information on the date and time of the onset of the disease caused by the pest or the like. Specifically. It includes information such as microbial disease, pest disease, physiological disorder, and physiological disorder caused by weeds. It also includes information on additional details of each type. For example, if it is a microbial disease, information on the type of disease, such as gray mold disease and leaf mold disease, is included.

In addition, the information on the occurrence of pests includes information on the incubation period from infection or parasitic to onset of the pest. The incubation period of a pest refers to the number of days required from infection to the onset of a pest, and the incubation period of a pest is when insects spawn, larvae hatch from eggs, and eat crops. It refers to the period until. That is, the incubation period refers to the period until damage to the crop occurs due to pests and diseases. This incubation period may be information obtained from a history of disease and pest occurrence, literature, or investigation.

In addition, the information processing terminal 200 may acquire not only information on occurrence of pests and pests, but also environmental information on a space in which crops are cultivated and environmental information (hereinafter, generally referred to as environmental information) about the cultivation status of crops. Further, the environmental information acquired by the information processing terminal 200 may be information that is difficult to detect by the cultivation sensor 300 described later.

As environmental information acquired by the information processing terminal 200, the cultivation situation of a crop can be mentioned, for example. The crop cultivation situation is specifically, the variety of the crop, the altitude of the crop cultivation site, the type of soil, the nutrient status of the soil, the sowing date and time of the crop, the planting date and time, and the situation of the red leaves of the leaves of the crop. , Information such as the history of application of the drug, the residual effect of the drug, the state of weeding, the state of water washing, the growth stage of the crop, etc. are included.

When environmental information is acquired by the information processing terminal 200, it may be performed by, for example, an auxiliary input operation by a user or the like. That is, environmental information may be acquired by the information processing system 1 not only by information acquired by the cultivation sensor 300 described later, but also by information acquired by the information processing terminal 200. The environmental information acquired by the information processing terminal 200 or the cultivation sensor 300 may be environmental information including a past history, or may be environmental information indicating a current value or a predicted value.

((Communication Department))

The communication unit 220 has a function of acquiring the pest occurrence information or environment information input from the input unit 210 and outputting the information to the information processing device 100. Further, the communication unit 220 also has a function of acquiring the information processed by the information processing device 100 and outputting it to the display unit 230. In the communication unit 220, when the pest occurrence information or environmental information is acquired, communication may be appropriately performed, or communication may be continuously performed. Further, communication may be performed in response to an instruction from the information processing device 100.

((Display))

The display unit 230 has a function of outputting information output from the communication unit 220. For example, the display unit 230 presents the infection probability predicted by the information processing device 100 to the user.

(Cultivation sensor)

The cultivation sensor 300 has a function of acquiring environmental information of a space in which crops are grown. The cultivation sensor 300 includes a sensor unit 310 that senses environmental information and a communication unit 320 that outputs the sensed environmental information.

((Sensor part))

The sensor unit 310 senses environmental information. The sensor unit 310 may be constituted by, for example, a thermometer, a hygrometer, an insolation meter, a carbon dioxide concentration meter, a moisture content meter that measures the moisture content of the soil, or the like.

The space in which crops of the environmental information acquired by the cultivation sensor 300 are cultivated may be a space surrounding a place where crops are cultivated, or may be an agricultural facility that cultivates crops such as a plastic house. If a place where crops are cultivated, it does not have to be surrounded, or an open space such as a field or rice field may be used.

The cultivation sensor 300 acquires environmental information such as temperature, humidity, insolation, carbon dioxide concentration, and moisture content of soil in which the crop is grown, for example, of the surrounding environment of the crop.

((Communication Department))

The communication unit 320 has a function of outputting the environmental information acquired by the sensor unit 310 to the information processing apparatus 100. In the communication unit 320, when the environmental information is acquired, communication may be appropriately performed, or communication may be continuously performed. Further, communication may be performed in response to an instruction from the information processing device 100.

(Weather server)

The meteorological server 400 has a function of outputting environmental information about the weather in a region where crops are cultivated. Examples of the environmental information to be output include temperature, humidity, solar radiation, rainfall, and the like. The meteorological server 400 may selectively output information used in the information processing device 100 by being requested from the information processing device 100. The environmental information output by the meteorological server 400 may be a past history, or may be a current value or a predicted value of the environmental information.

(Information processing device)

The information processing device 100 acquires the information output by the information processing terminal 200, the cultivation sensor 300, and the meteorological server 400, and an infection that leads to the outbreak of diseases such as diseases on crops caused by pests has occurred. It mainly has the function of estimating the time of infection. The information processing apparatus 100 includes an acquisition unit 110, a storage unit 120, an estimated infection period determination unit 130, an infection occurrence estimation unit 140, a machine learning unit 150, and an infection. It includes an occurrence prediction unit 160.

((Acquisition Department))

The acquisition unit 110 has a function of acquiring environmental information and disease occurrence information output by the information processing terminal 200, the cultivation sensor 300, and the meteorological server 400. The acquired environmental information and disease occurrence information are stored in the storage unit 120.

((Memory))

The storage unit 120 has a function of storing information acquired by the acquisition unit 110. Specifically, the storage unit 120 stores the disease occurrence information and environmental information acquired by the acquisition unit 110. Further, the storage unit 120 also stores parameters related to machine learning in infection estimation described later, an infection prediction model, and the like.

((Presumed infection period decision part))

The estimated infection period determining unit 130 uses the information on the occurrence of pests to determine the estimated infection period suspected of having an infection leading to the outbreak of the crop and the no-infection period in which the infection leading to the outbreak has not occurred. Have.

The estimated infection period may be a period including at least one point during which an infection leading to an outbreak to crops occurred. Specifically, the estimated infection period may include the date and time at which the infection to the crop occurred. In addition, the estimated infection period may include a plurality of periods during which infection to crops occurs.

In the estimated infection period, the date and time at which the infection occurred and the date and time at which the infection did not occur are mixed and included. As the estimated infection period, for example, the estimated infection period determining unit 130 may determine, as the estimated infection period, a period at the time of an outbreak, which has moved back from the number of days of the incubation period of the pest. The estimated infection period should just include the period in which infection occurs, and may be appropriately determined according to the type of pest and the necessary precision of infection prediction. For example, when there is a possibility that the incubation period of a known pest is likely to change depending on the environment or the like, the incubation period may be set longer and the estimated infection period may be set longer.

On the other hand, in the estimated infection period determining unit 130, a period in which infection has not occurred may be determined as the period without infection. For example, from a point in time when no outbreak has been confirmed, a period of at least a period back from the incubation period of the pest is determined as the period of no infection. The period of no infection may be a period including a continuous date and time of no infection, or may be a period with an arbitrary date and time as a time point. In addition, as the infection-free period, a plurality of date and time may be determined with an arbitrary date and time as one time point.

((Infection Occurrence Estimation Department))

The infection occurrence estimation unit 140 compares the environmental information of the estimated infection period determined by the estimated infection period determination unit 130 with the environmental information of the no-infection period, and estimates the time of infection at which the infection occurred from the estimated infection period. Has a function.

With reference to FIG. 4, the estimation of infection by the infection occurrence estimation unit 140 will be described in detail. In Fig. 4, the time axis T is shown. Here, first, the onset point A is set to November 30. When the incubation period of the pest is between 3 and 7 days, the start point C of the estimated infection period is November 23, which is the longest number of days of the incubation period from the onset date A. In addition, the end point B of the estimated infection period is November 27, which is the shortest number of days of the incubation period from the onset date and time A.

On the other hand, the infection-free period may be appropriately determined as long as it is a known date and time that no infection has occurred, and is at least a period before the start point C of the estimated infection period. In addition, the period without infection may be a period other than the estimated infection period. In Fig. 4, the no-infection period is determined as the date and time D, that is, October 21, and the date and time E, that is, October 20, before the start point C of the estimated infection period.

The infection occurrence estimation unit 140 designates and compares the environmental information of the estimated infection period and the environmental information of the non-infection period. Environmental information of the estimated infection period is designated as PositiveBag (PB), and environmental information of the no-infection period is designated as NegativeBag (NB), respectively. In addition, the estimated infection period is 5 days from November 23 to November 27.

Here, environmental information corresponding to the date and time classified as PB is designated as PB1, November 24 as PB2, November 25 as PB3, and November 26 as PB4. , November 27th is designated as PB5. On the other hand, environment information corresponding to the date and time classified as NB is designated as NB1 on October 20 and NB2 on October 21, respectively.

Factors included in environmental information are schematically represented by white circles, black circles, black triangles, and white triangles. Environmental information NB1 of date and time E on October 20, when the infection-free period, includes 4 white circles, 2 black circles, and 2 black triangles, and environmental information NB2 on date and time D on October 21 Contains 4 white circles, 3 black circles, and 2 black triangles. On the other hand, at the date of the estimated infection period, environmental information PB1 on November 23 contains 4 white circles, 2 black circles, 2 black triangles, and 3 white triangles. The environmental information of work PB2 contains four white circles, three black circles, one black triangle, and one white triangle. Similarly, environmental information PB3 on November 25 includes 3 white circles, 3 black circles, and 2 black triangles, and environmental information PB4 on November 26 contains 4 white circles and black circles. These 2, 2 black triangles, 3 white triangles are included, and environmental information PB5 of November 27 contains 3 white circles, 3 black circles, 2 black triangles, and 1 white triangle. Included.

Here, the infection occurrence estimation unit 140 compares the environmental information NB1 and the environmental information NB2 during the no-infection period with the environmental information PB1 to the environmental information PB5 during the estimated infection period. In the infection occurrence estimation unit 140, by this comparison, the white triangle is not included in the environmental information NB1 and the environmental information NB2 of the no-infection period, and the white triangle is not included in the environmental information PB1 to the environmental information PB5 of the estimated infection period. Know what's included. As a result, the white triangle is extracted as a feature quantity S leading to an outbreak among the factors of environmental information.

The infection occurrence estimation unit 140 further calculates an infection probability due to a change in the characteristic quantity S during the estimated infection period using the extracted characteristic quantity S. In Fig. 4, the date and time X of the no-infection period and the estimated infection period, the number of days Y back from the onset of the crop to each date, and the change Z of the characteristic amount at each date are shown as a table. In addition, specifically, in FIG. 4, the characteristic quantity S is shown as humidity (%).

Paying attention to the column Z of the change in feature quantity, the humidity (%) is 95% on November 23, 60% on November 24, 55% on November 25, 95% on November 26, and 50% on November 27. to be. On the other hand, the humidity in the infection-free period is 55% on October 20 and 60% on October 21. Using this feature amount S, the infection occurrence estimating unit 140 estimates the time of infection from during the estimated infection period.

For example, in the example shown in FIG. 4, since the higher the humidity, the higher the infection probability, the estimated infection period, November 23 and November 26, when the humidity was the highest value, are estimated as the time of infection. In this example, the higher the value of the feature, the higher the probability of infection, but the lower the value of the feature, the higher the probability of infection may be.

In addition, the probability of occurrence of infection may be calculated according to the type of the characteristic quantity and the change trend of the characteristic quantity, for example, the more rapid the change in the characteristic quantity, the higher the probability of occurrence of infection, and so on. In addition, in the infection occurrence estimation unit 140, in the estimated infection period, a plurality of infection dates may be estimated, or one date and time may be estimated.

The infection occurrence estimating unit 140 may set a threshold value to the infection probability and determine whether or not an infection has occurred. For example, when an infection probability higher than the threshold value is shown, it may be estimated as the time of infection. On the other hand, when an infection probability lower than the threshold value is shown, the infection occurrence estimation unit 140 may calculate the infection probability again.

In addition, the infection occurrence estimation unit 140 may have a function of determining whether the estimated infection time is appropriate and selecting an appropriate infection time in order to increase the reliability of the estimated infection time. For example, during the estimated infection period, when a plurality of infection dates are estimated, it may be determined that the estimated infection time is not appropriate, and a more reliable infection time may be selected from the plurality of infection times. This function may be performed by the user. That is, it is determined whether or not the estimated plurality of infection times are appropriate by the user, and an appropriate infection time may be selected by the user from among the estimated plurality of infection times. In this way, the infection occurrence estimation unit 140 estimates the infection time more precisely.

((Machine Learning Department))

The machine learning unit 150 performs machine learning on environmental information at the time of infection estimated by the infection occurrence estimation unit 140, and has a function of constructing an infection prediction model for predicting a current or future infection probability. Have. In the infection occurrence estimating unit 140, since the time of infection is estimated, information such as which factor in the environmental information causes the infection is accumulated with respect to the infection that will lead to the outbreak. By machine learning the environmental information that causes the infection, the machine learning unit 150 can construct an infection prediction model showing the relationship between each factor in the environmental information and the infection.

The machine learning unit 150 constructs an infection prediction model and additionally outputs it to the storage unit 120. The infection prediction model stored in the storage unit 120 is appropriately used for predicting the infection of crops.

((Infection occurrence prediction unit))

The infection occurrence prediction unit 160 has a function of predicting an infection probability using the infection prediction model stored in the storage unit 120 and the current value or predicted value of environmental information of the crop. Here, the environmental information of the crop is environmental information on the crop for which the outbreak of pests is predicted, and is output by at least one of the environmental information output from the cultivation sensor 300 and the environmental information output from the meteorological server 400 It may be environmental information.

The present value or the predicted value represents current environmental information or predicted environmental information about the crop. For example, the present value may be environmental information continuously acquired by the cultivation sensor 300 up to now, or may be environmental information periodically updated and acquired. In addition, the current value may be appropriately acquired by the information processing apparatus 100 or the instruction of the user, etc., even if it is not updated regularly. In addition, the predicted value indicates future environmental information after the present value, and indicates predicted values such as several hours, days, and weeks from the present. For example, the predicted value of environmental information represents a weather forecast or the like output from the meteorological server 400.

The infection occurrence prediction unit 160 predicts the probability of occurrence of infection using the current value or predicted value of the above-described environmental information. The infection probability may be an infection probability at an arbitrary time point with respect to a change over time of a crop, or may be a successive infection probability corresponding to a change over time. It is also possible to predict not only the probability but also whether or not infection will occur. In addition, the probability of infection may be predicted for the past.

In addition, the infection occurrence prediction unit 160 may calculate an infection probability for each type of pest. Thereby, it is possible to predict the probability of infection with respect to a plurality of diseases and pests, and reinforcement of measures against diseases and pests is possible.

The infection occurrence prediction unit 160 also has a function of outputting the predicted infection probability to the information processing terminal 200. In the information processing terminal 200, for example, the infection probability is displayed on the display unit 230 and presented to the user. The user can check the infection probability at the information processing terminal 200 to plan spraying of drugs such as pesticides. In addition, the output destination of the infection probability is not limited to the information processing terminal 200, and may be a plurality of information processing terminals.

<3. Operation flow>

So far, the functional configuration of the information processing system 1 has been described. Next, an operation flow of the information processing system 1 will be described with reference to FIGS. 5 to 7.

In the information processing system 1, roughly two processes are performed. That is, the information processing system 1 first constructs an infection prediction model (S100), and then uses the constructed infection prediction model to predict an infection probability (S200). Hereinafter, each operation flow will be described in detail.

(Establishment of infection prediction model)

When constructing an infection prediction model, first, in the acquisition unit 110 of the information processing device 100, the information on the occurrence of pests and the environment is acquired (S102). The pest occurrence information includes information indicating the onset date and time, and is acquired by the information processing device 100 through the information processing terminal 200 or the like. The environmental information is, for example, information output from the cultivation sensor 300 installed in the cultivation environment of the crop or the meteorological server 400, and information processing through at least one of the cultivation sensor 300 or the meteorological server 400 It can be acquired in the device 100.

Next, in the estimated infection period determining unit 130 of the information processing apparatus 100, the estimated infection period and the non-infection period are determined (S104). The estimated infection period is a period including a period in which at least one disease infection of the crop occurred, and the no-infection period is a period in which no infection has been confirmed.

Here, in the estimated infection period determining unit 130, for example, a Bag number is assigned to a plurality of environmental information in the estimated infection period and the non-infection period, and the environmental information corresponding to the estimated infection period is assigned a PositiveBag (PB ) Number, and NegativeBag(NB) number to environmental information corresponding to the period of no infection.

When selecting the environmental information NB corresponding to the infection-free period, the NB number is assigned to the environmental information corresponding to the infection-free period from the history of the disease occurrence information in which the outbreak of the pest, which is the target of the construction of the infection prediction model, is confirmed. At this time, by using the pest occurrence information reported by a plurality of farms as the pest occurrence information, the number of NBs can be increased to increase the precision of the infection prediction model. For example, NB1 is assigned to environmental information in an infection-free period based on disease occurrence information in one farm, and NB2 is assigned to environmental information in an infection-free period based on disease occurrence information in another farm. In this way, the NB number is assigned to the environmental information of the infection-free period corresponding to the infection by the crop predicting the infection and the crop of the same kind as the crop and the pest of the same type, and by designating the NB number, the following In the treatment, the time of infection is estimated.

Next, the infection occurrence estimation unit 140 compares the environmental information of the estimated infection period and the environmental information of the non-infection period, and estimates the time of infection during the estimated infection period (S106). By comparing the environmental information to which the above-described PB number and NB number are assigned, a feature amount at which infection occurs is extracted, and the estimation at the time of infection is performed.

Further, the infection occurrence estimation unit 140 determines whether or not the estimated infection time has been properly estimated (S108). For example, this determination may be made by the user confirming the estimated infection time.

If the estimated infection time is appropriate (S108/Yes), the process proceeds to the next process. On the other hand, if the estimated time of infection is inappropriate (S108/No), the infection occurrence estimation unit 140 compares the environmental information of the estimated infection period and the environmental information of the no-infection period, and estimates the time of infection. good.

In addition, that the estimated infection time is inappropriate is when there are a plurality of estimated infection times within the estimated infection period, or when the probability of infection is an appropriate range in the estimation of the infection time, and a threshold range is set, The case of the infection probability deviating from the threshold value range, etc. are mentioned.

Next, the infection occurrence estimation unit 140 performs a determination process at the time of the estimated infection (S110). If the infection time estimated in the pretreatment is appropriate (S108/Yes), the determination process at the time of infection is performed as it is. On the other hand, if the estimated time of infection is inappropriate (S108/No), the user or the like selects the time of infection with the highest certainty of infection from the estimated time of infection without re-estimating the time of infection, and selects the time of infection. You may decide.

Next, the machine learning unit 150 performs machine learning on the environment information at the time of infection in the estimation of the time of infection, and constructs an infection prediction model (S112).

(Infection prediction using infection prediction model)

So far, the operation flow until constructing an infection prediction model has been described. Next, with reference to FIG. 7, the prediction of the infection of the crop by using the above infection prediction model will be described.

First, the infection occurrence prediction unit 160 acquires an infection prediction model from the storage unit 120 (S202). In the storage unit 120, an infection prediction model constructed by the above-described method is stored, and an infection prediction model is output from the storage unit 120 in response to an instruction of the infection occurrence prediction unit 160.

In addition, the infection occurrence prediction unit 160 acquires a current value or a predicted value of the environmental information stored in the storage unit 120 (S204). The storage unit 120 stores a current value or a predicted value of environmental information acquired from the cultivation sensor 300 or the meteorological server 400, and according to an instruction of the infection occurrence prediction unit 160, the storage unit 120 The current value or predicted value of environmental information is output from.

Next, the infection occurrence prediction unit 160 predicts the infection probability using the infection prediction model and the current value or predicted value of the environmental information of the crop (S206).

The infection occurrence prediction unit 160 outputs the predicted infection probability (S208). The output destination may be, for example, a terminal possessed by a user who needs an infection probability such as the information processing terminal 200, or a terminal such as a computer or a smartphone.

In the information processing system 1 according to the above-described operation flow, the estimation at the time of infection is accurately performed. In addition, it is also possible to predict the probability of future infections for crops.

<4. Hardware configuration example>

Next, with reference to Fig. 8, the hardware configuration of the information processing apparatus according to the present embodiment will be described. 8 is a block diagram showing an example of a hardware configuration of the information processing system 1 according to the present embodiment. In addition, the information processing apparatus 900 shown in FIG. 8 can implement the information processing system 1 shown in FIG. 3, for example. Information processing by the information processing system 1 according to the present embodiment is realized by cooperation between software and hardware described below.

As shown in Fig. 8, the information processing device 900 includes a CPU (Central Processing Unit) 901, a ROM (Read Only Memory) 902, a RAM (Random Access Memory) 903, and a host bus 904a. It is equipped with. In addition, the information processing device 900 includes a bridge 904, an external bus 904b, an interface 905, an input device 906, an output device 907, a storage device 908, a drive 909, and a connection. It has a port 911 and a communication device 913. The information processing device 900 may have a processing circuit such as an electric circuit, a digital signal processor (DSP) or an application specific integrated circuit (ASIC) in place of or in addition to the CPU 901.

The CPU 901 functions as an arithmetic processing device and a control device, and controls overall operations in the information processing device 900 according to various programs. Further, the CPU 901 may be a microprocessor. The ROM 902 stores programs, operation parameters, and the like used by the CPU 901. The RAM 903 temporarily stores programs used in the execution of the CPU 901, parameters that change appropriately during the execution, and the like. The CPU 901 can execute the functions of the information processing device 100 or the information processing terminal 200 shown in FIG. 3, for example.

The CPU 901, ROM 902, and RAM 903 are interconnected by a host bus 904a including a CPU bus or the like. The host bus 904a is connected via a bridge 904 to an external bus 904b such as a PCI (Peripheral Component Interconnect/Interface) bus. In addition, it is not always necessary to separate the host bus 904a, the bridge 904, and the external bus 904b, and such a function may be mounted on one bus.

The input device 906 is realized by a device to which information is input by a user, such as a mouse, a keyboard, a touch panel, a button, a microphone, a switch, and a lever, for example. Further, the input device 906 may be, for example, a remote control device using infrared rays or other radio waves, and an external connection device such as a mobile phone or PDA (Personal Digital Assistant) corresponding to the operation of the information processing device 900 It can be done. In addition, the input device 906 includes, for example, an input control circuit that generates an input signal based on information input by the user using the above input means, and outputs it to the CPU 901. You may have it. A user of the information processing device 900 can input various types of data or instruct a processing operation to the information processing device 900 by operating the input device 906. The input device 906 may correspond to the input unit 210 of the information processing terminal 200 shown in FIG. 3.

The output device 907 is formed of a device capable of notifying a user of the acquired information visually or aurally. As such devices, display devices such as CRT (Cathode Ray Tube) display devices, liquid crystal display devices, plasma display devices, EL (electroluminescence) display devices, laser projectors, LED projectors and lamps, and audio output devices such as speakers and headphones, etc. There is this. The output device 907 outputs, for example, results obtained by various processes performed by the information processing device 900. Specifically, the output device 907 visually displays results obtained by various processes performed by the information processing device 900 in various formats such as text, images, tables, and graphs. On the other hand, in the case of using an audio output device, an audio signal composed of reproduced audio data or audio data is converted into an analog signal and audibly output. The output device 907 can execute, for example, a function of the display unit 230 of the information processing terminal 200 shown in FIG. 3.

The storage device 908 is a device for storing data formed as an example of a storage unit of the information processing device 900. The storage device 908 is realized by, for example, a magnetic storage device such as a hard disk drive (HDD), a semiconductor storage device, an optical storage device, or a magneto-optical storage device. The storage device 908 may include a storage medium, a recording device for writing data to the storage medium, a reading device for reading data from the storage medium, and a deletion device for deleting data recorded on the storage medium. The storage device 908 stores programs executed by the CPU 901, various data, and various data acquired from the outside. The storage device 908 can execute, for example, a function of the storage unit 120 of the information processing device 100 shown in FIG. 3.

The drive 909 is a storage medium reader/writer, and is built-in or external to the information processing device 900. The drive 909 reads information recorded on a removable storage medium such as a mounted magnetic disk, optical disk, magneto optical disk, or semiconductor memory, and outputs it to the RAM 903. Further, the drive 909 can also record information on a removable storage medium.

The connection port 911 is an interface to be connected to an external device, and is a connection port to an external device capable of transmitting data by, for example, a Universal Serial Bus (USB) or the like.

The communication device 913 is, for example, a communication interface formed of a communication device or the like for connecting to the network 920. The communication device 913 is, for example, a communication card for wired or wireless LAN (Local Area Network), LTE (Long Term Evolution), Bluetooth (registered trademark) or WUSB (Wireless USB). Further, the communication device 913 may be a router for optical communication, a router for ADSL (Asymmetric Digital Subscriber Line), or a modem for various communication. The communication device 913 can transmit and receive signals and the like between, for example, the Internet or other communication devices according to a predetermined protocol such as TCP/IP. The communication device 913 can execute functions such as the communication unit 220 of the cultivation sensor 300 shown in FIG. 3 or the communication unit 220 of the information processing terminal 200, for example.

Further, the network 920 is a wired or wireless transmission path of information transmitted from a device connected to the network 920. For example, the network 920 may include a public line network such as the Internet, a telephone line network, and a satellite communication network, various local area networks (LANs) including Ethernet (registered trademarks), and wide area networks (WANs). Further, the network 920 may include a dedicated line network such as an Internet Protocol-Virtual Private Network (IP-VPN).

In addition, a computer program for exerting functions equivalent to the respective configurations of the information processing system 1 according to the present embodiment described above can be created by hardware such as CPU, ROM, and RAM built into the information processing system 1. Further, a recording medium in which the computer program is stored is also provided.

<5. Variation example>

So far, when the present invention estimates the occurrence of events leading to the occurrence of damage to crops, the occurrence of diseases and pests has been mainly described as an example. The present invention is not limited to this example, and can be applied to a case in which the occurrence of damage and the occurrence of events leading to the damage do not occur at the same time. Specifically, the present invention can also be used for infectious diseases in animals, food poisoning in animals, and the like.

(Animal infectious disease)

In the present embodiment, an example in which a disease or a pest occurs in a crop was described, but the present invention is applicable to the identification of the date and time of infection when an animal develops an infectious disease. Until the animal develops an infectious disease, the infection of the infectious disease occurs, and the infectious disease develops through the incubation period of the infectious disease. In the case of infectious diseases of animals, more often than after the onset of the infectious disease, it is more effective to take a drug before onset, particularly before and after infection, to have a higher infectious disease suppression effect. Therefore, in growing animals, it is important to predict the probability of infection.

In the case of animal infectious diseases, when the present invention is applied, infectious disease occurrence information including the time of occurrence of damage to the animal, that is, the occurrence date and time of the infectious disease, and environmental information including the growth environment of the animal are acquired. Using the infectious disease incidence information, the period in which the animal is suspected of being infected with an infectious disease is assumed to be the period of occurrence of the event, and the period in which there is no suspicion of infection as the period of no death.

By comparing the estimated event occurrence period and the growth environment information of the non-death period, a feature amount that becomes a factor infecting an infectious disease is extracted from the growth environment information. Using this feature quantity, the event occurrence time is estimated from the estimated event occurrence period.

(Food poisoning in animals)

The present invention can also be applied to the identification of the causative food or the timing of ingestion of poison when an animal develops food poisoning. Until the animal develops food poisoning, the ingestion of food poisoning and the growth of viruses or bacteria that cause food poisoning occur. When the proliferation exceeds the threshold, symptoms of food poisoning develop in the animal. In the case of food poisoning, there is a high possibility that symptoms of food poisoning can be alleviated or suppressed by taking a drug before the proliferation of viruses or bacteria occurs. In addition, since the drug to be taken varies depending on the cause of food poisoning, it becomes important to alleviate the symptoms that the cause of food poisoning is specified.

In the case of food poisoning of animals, when the present invention is applied, the occurrence of food poisoning is the occurrence of damage. In other words, information on occurrence of food poisoning including the date and time of onset of food poisoning, and information on ingestion by which the animal has consumed food or the like is obtained. Using this information on occurrence of food poisoning, the period of suspected ingestion of food, etc. that causes food poisoning is assumed to be the period of occurrence of events, and the period of no suspicion of ingestion is the period of uneventfulness.

By comparing the estimated event occurrence period with the intake information during the uneventful period, a characteristic amount that causes food poisoning, that is, food, etc., is extracted from the intake information. Using this feature quantity, it is possible to estimate the event occurrence time from the estimated event occurrence period.

In the above, preferred embodiments of the present invention have been described in detail with reference to the accompanying drawings, but the present invention is not limited to these examples. It is clear that those of ordinary skill in the field of technology to which the present invention belongs can conceive of various modifications or modifications within the scope of the technical idea described in the claims, and also for these, Naturally, it is understood to be within the technical scope of the present invention.

100 information processing unit
110 Acquisition Department
120 memory
130 Estimated Infection Period Decision Part
140 Infection Occurrence Estimator
150 Machine Learning Department
160 Infection occurrence prediction unit
200 information processing terminals
210 input
220 Communication Department
230 display
300 cultivation sensor
310 Sensor part
320 Communication Department
400 weather servers

Claims (9)

An acquisition unit that acquires pest occurrence information including information at the time of damage to the crop caused by pests and environmental information including the cultivation environment of the crop;
Using the information on the occurrence of the disease and pests, determining the estimated event occurrence period in which the event leading to the occurrence of the damage to the crop has occurred and the period in which the event leading to the occurrence of the damage has not occurred is determined. Wealth,
An information processing apparatus comprising: an event estimating unit for comparing the environment information of the estimated event occurrence period with the environment information of the non-event period, and estimating the occurrence time of the event in which the event has occurred from the estimated event occurrence period. The information processing apparatus according to claim 1, wherein the event estimating unit extracts a feature amount that affects the occurrence of the damage from among the environmental information during the estimated event occurrence period. The method of claim 2, wherein the event occurrence probability of the event occurring in the estimated event occurrence period is calculated using the feature quantity
The information processing device for estimating the occurrence time of the event based on the occurrence probability of the event. The information processing apparatus according to any one of claims 1 to 3, wherein the estimated event occurrence period is a period prior to the occurrence of the damage to the crop, and at least one period including when the event occurs is included. . The method of claim 3, further comprising a machine learning unit configured to construct an event occurrence prediction model for predicting the event occurrence probability by performing machine learning on the environmental information at the time of the event occurrence estimated by the event estimation unit. , Information processing device. The information processing apparatus according to claim 5, wherein the event occurrence prediction model is used to predict the occurrence probability of the event of the crop using the current value or the predicted value of the environmental information of the crop. The information processing apparatus according to claim 6, further comprising an output unit that outputs the probability of occurrence of the event according to the type of the pest. Computer,
An acquisition unit that acquires pest occurrence information including information at the time of damage to the crop caused by pests and environmental information including the cultivation environment of the crop;
Using the information on the occurrence of the disease and pests, determining the estimated event occurrence period in which the event leading to the occurrence of the damage to the crop has occurred and the period in which the event leading to the occurrence of the damage has not occurred is determined. Wealth,
A program for functioning as an event estimation unit, which compares the environmental information of the estimated event occurrence period with the environmental information during the non-event period, and estimates the occurrence time of the event in which the event occurs from the estimated event occurrence period. An acquisition unit that acquires pest occurrence information including information at the time of damage to the crop caused by pests and environmental information including the cultivation environment of the crop;
Using the information on the occurrence of the disease and pests, determining the estimated event occurrence period in which the event leading to the occurrence of the damage to the crop has occurred and the period in which the event leading to the occurrence of the damage has not occurred is determined. Wealth,
And an event estimation unit for comparing the environment information of the estimated event occurrence period with the environment information of the non-event period, and estimating when the event occurs from the estimated event occurrence period.

Images